1. 14 Dec, 2018 8 commits
    • Logan Gunthorpe's avatar
      mm: introduce common STRUCT_PAGE_MAX_SHIFT define · d1402fc7
      Logan Gunthorpe authored
      This define is used by arm64 to calculate the size of the vmemmap
      region.  It is defined as the log2 of the upper bound on the size of a
      struct page.
      
      We move it into mm_types.h so it can be defined properly instead of set
      and checked with a build bug.  This also allows us to use the same
      define for riscv.
      
      Link: http://lkml.kernel.org/r/20181107205433.3875-2-logang@deltatee.comSigned-off-by: default avatarLogan Gunthorpe <logang@deltatee.com>
      Acked-by: default avatarWill Deacon <will.deacon@arm.com>
      Acked-by: default avatarAndrew Morton <akpm@linux-foundation.org>
      Acked-by: default avatarArd Biesheuvel <ard.biesheuvel@linaro.org>
      Acked-by: default avatarCatalin Marinas <catalin.marinas@arm.com>
      Cc: Arnd Bergmann <arnd@arndb.de>
      Cc: Christoph Hellwig <hch@lst.de>
      Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
      Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      d1402fc7
    • Mike Rapoport's avatar
      alpha: fix hang caused by the bootmem removal · 5b526090
      Mike Rapoport authored
      The conversion of alpha to memblock as the early memory manager caused
      boot to hang as described at [1].
      
      The issue is caused because for CONFIG_DISCTONTIGMEM=y case,
      memblock_add() is called using memory start PFN that had been rounded
      down to the nearest 8Mb and it caused memblock to see more memory that
      is actually present in the system.
      
      Besides, memblock allocates memory from high addresses while bootmem was
      using low memory, which broke the assumption that early allocations are
      always accessible by the hardware.
      
      This patch ensures that memblock_add() is using the correct PFN for the
      memory start and forces memblock to use bottom-up allocations.
      
      [1] https://lkml.org/lkml/2018/11/22/1032
      
      Link: http://lkml.kernel.org/r/1543233216-25833-1-git-send-email-rppt@linux.ibm.comReported-by: default avatarMeelis Roos <mroos@linux.ee>
      Signed-off-by: default avatarMike Rapoport <rppt@linux.ibm.com>
      Tested-by: default avatarMeelis Roos <mroos@linux.ee>
      Cc: Richard Henderson <rth@twiddle.net>
      Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
      Cc: Matt Turner <mattst88@gmail.com>
      Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
      Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      5b526090
    • Linus Torvalds's avatar
      Merge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux · eb6cf9f8
      Linus Torvalds authored
      Pull arm64 fix from Catalin Marinas:
       "Invalidate the caches before clearing the DMA buffer via the
        non-cacheable alias in the FORCE_CONTIGUOUS case"
      
      * tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
        arm64: dma-mapping: Fix FORCE_CONTIGUOUS buffer clearing
      eb6cf9f8
    • Linus Torvalds's avatar
      Merge tag 'powerpc-4.20-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux · 4645453c
      Linus Torvalds authored
      Pull powerpc fixes from Michael Ellerman:
       "One notable fix for our change to split pt_regs between user/kernel,
        we forgot to update BPF to use the user-visible type which was an ABI
        break for BPF programs.
      
        A slightly ugly but minimal fix to do_syscall_trace_enter() so that we
        use tracehook_report_syscall_entry() properly. We'll rework the code
        in next to avoid the empty if body.
      
        Seven commits fixing bugs in the new papr_scm (Storage Class Memory)
        driver. The driver was finally able to be tested on the other
        hypervisor which exposed several bugs. The fixes are all fairly
        minimal at least.
      
        Fix a crash in our MSI code if an MSI-capable device is plugged into a
        non-MSI capable PHB, only seen on older hardware (MPC8378).
      
        Fix our legacy serial code to look for "stdout-path" since the device
        trees were updated to use that instead of "linux,stdout-path".
      
        A change to the COFF zImage code to fix booting old powermacs.
      
        A couple of minor build fixes.
      
        Thanks to: Benjamin Herrenschmidt, Daniel Axtens, Dmitry V. Levin,
        Elvira Khabirova, Oliver O'Halloran, Paul Mackerras, Radu Rendec, Rob
        Herring, Sandipan Das"
      
      * tag 'powerpc-4.20-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
        powerpc/ptrace: replace ptrace_report_syscall() with a tracehook call
        powerpc/mm: Fallback to RAM if the altmap is unusable
        powerpc/papr_scm: Use ibm,unit-guid as the iset cookie
        powerpc/papr_scm: Fix DIMM device registration race
        powerpc/papr_scm: Remove endian conversions
        powerpc/papr_scm: Update DT properties
        powerpc/papr_scm: Fix resource end address
        powerpc/papr_scm: Use depend instead of select
        powerpc/bpf: Fix broken uapi for BPF_PROG_TYPE_PERF_EVENT
        powerpc/boot: Fix build failures with -j 1
        powerpc: Look for "stdout-path" when setting up legacy consoles
        powerpc/msi: Fix NULL pointer access in teardown code
        powerpc/mm: Fix linux page tables build with some configs
        powerpc: Fix COFF zImage booting on old powermacs
      4645453c
    • Linus Torvalds's avatar
      Merge tag 'ceph-for-4.20-rc7' of https://github.com/ceph/ceph-client · c19bf74f
      Linus Torvalds authored
      Pull ceph fix from Ilya Dryomov:
       "Luis discovered a problem with the new copyfrom offload on the server
        side. Disable it for now"
      
      * tag 'ceph-for-4.20-rc7' of https://github.com/ceph/ceph-client:
        ceph: make 'nocopyfrom' a default mount option
      c19bf74f
    • Linus Torvalds's avatar
      Merge tag 'pinctrl-v4.20-3' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl · 7fb5f5d3
      Linus Torvalds authored
      Pull pin control fixes from Linus Walleij:
       "Three pin control fixes for the v4.20 series. Just odd drivers, so
        nothing particularly interesting:
      
         - Set the tile property on Qualcomm SDM60.
      
         - Fix up enable register calculation for the Meson
      
         - Fix an IRQ offset on the Sunxi (Allwinner)"
      
      * tag 'pinctrl-v4.20-3' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
        pinctrl: sunxi: a83t: Fix IRQ offset typo for PH11
        pinctrl: meson: fix pull enable register calculation
        pinctrl: sdm660: Set tile property for pingroups
      7fb5f5d3
    • Linus Torvalds's avatar
      Merge tag 'drm-fixes-2018-12-14' of git://anongit.freedesktop.org/drm/drm · 92de1de5
      Linus Torvalds authored
      Pull drm fixes from Dave Airlie:
       "While I hoped things would calm down, the world hasn't joined with me,
        but it's a few things scattered over a wide area. The i915 workarounds
        regression fix is probably the largest, the rest are more usual sized.
        We also get some new AMD PCI IDs.
      
        There is also a patch in here to MAINTAINERS to added Daniel as an
        official DRM toplevel co-maintainer, he's decided he wants to step up
        and share the glory, and he'll likely process next weeks fixes while
        I'm away on holidays.
      
        Summary:
      
        amdgpu:
      
         - some new PCI IDs
      
         - fixed firmware image updates
      
         - power management fixes
      
         - locking warning fix
      
        nouveau:
      
         - framebuffer flushing fix
      
         - memory leak fix
      
         - tegra device init regression fix
      
        vmwgfx:
      
         - OOM kernel memory fix
      
         - excess return in function fix
      
        i915:
      
         - the biggest fix is a regression fix where workarounds weren't
           getting reapplied after a gpu hang causing further crashing, this
           fixes the workaround application to make it happen again
      
         - GPU hang fixes for Braswell and some GEN3 GPUs
      
         - GVT fix for broadwell tiling
      
        rockchip:
      
         - revert to fix a regression causing a WARN on shutdown
      
        mediatek:
      
         - avoid crash attaching to non-existant bridges"
      
      * tag 'drm-fixes-2018-12-14' of git://anongit.freedesktop.org/drm/drm: (23 commits)
        drm/vmwgfx: Protect from excessive execbuf kernel memory allocations v3
        MAINTAINERS: Daniel for drm co-maintainer
        drm/amdgpu: drop fclk/gfxclk ratio setting
        drm/vmwgfx: remove redundant return ret statement
        drm/i915: Flush GPU relocs harder for gen3
        drm/i915: Allocate a common scratch page
        drm/i915/execlists: Apply a full mb before execution for Braswell
        drm/nouveau/kms: Fix memory leak in nv50_mstm_del()
        drm/nouveau/kms/nv50-: also flush fb writes when rewinding push buffer
        drm/amdgpu: Fix DEBUG_LOCKS_WARN_ON(depth <= 0) in amdgpu_ctx.lock
        Revert "drm/rockchip: Allow driver to be shutdown on reboot/kexec"
        drm/nouveau/drm/nouveau: tegra: Call nouveau_drm_device_init()
        drm/amdgpu/powerplay: Apply avfs cks-off voltages on VI
        drm/amdgpu: update SMC firmware image for polaris10 variants
        drm/amdkfd: add new vega20 pci id
        drm/amdkfd: add new vega10 pci ids
        drm/amdgpu: add some additional vega20 pci ids
        drm/amdgpu: add some additional vega10 pci ids
        drm/amdgpu: update smu firmware images for VI variants (v2)
        drm/i915: Introduce per-engine workarounds
        ...
      92de1de5
    • Linus Torvalds's avatar
      Merge tag 'xarray-4.20-rc7' of git://git.infradead.org/users/willy/linux-dax · 880b9df1
      Linus Torvalds authored
      Pull XArray fixes from Matthew Wilcox:
       "Two bugfixes, each with test-suite updates, two improvements to the
        test-suite without associated bugs, and one patch adding a missing
        API"
      
      * tag 'xarray-4.20-rc7' of git://git.infradead.org/users/willy/linux-dax:
        XArray: Fix xa_alloc when id exceeds max
        XArray tests: Check iterating over multiorder entries
        XArray tests: Handle larger indices more elegantly
        XArray: Add xa_cmpxchg_irq and xa_cmpxchg_bh
        radix tree: Don't return retry entries from lookup
      880b9df1
  2. 13 Dec, 2018 13 commits
  3. 12 Dec, 2018 17 commits
  4. 11 Dec, 2018 2 commits
    • Chad Austin's avatar
      fuse: continue to send FUSE_RELEASEDIR when FUSE_OPEN returns ENOSYS · 2e64ff15
      Chad Austin authored
      When FUSE_OPEN returns ENOSYS, the no_open bit is set on the connection.
      
      Because the FUSE_RELEASE and FUSE_RELEASEDIR paths share code, this
      incorrectly caused the FUSE_RELEASEDIR request to be dropped and never sent
      to userspace.
      
      Pass an isdir bool to distinguish between FUSE_RELEASE and FUSE_RELEASEDIR
      inside of fuse_file_put.
      
      Fixes: 7678ac50 ("fuse: support clients that don't implement 'open'")
      Cc: <stable@vger.kernel.org> # v3.14
      Signed-off-by: default avatarChad Austin <chadaustin@fb.com>
      Signed-off-by: default avatarMiklos Szeredi <mszeredi@redhat.com>
      2e64ff15
    • Mike Snitzer's avatar
      dm thin: send event about thin-pool state change _after_ making it · f6c36758
      Mike Snitzer authored
      Sending a DM event before a thin-pool state change is about to happen is
      a bug.  It wasn't realized until it became clear that userspace response
      to the event raced with the actual state change that the event was
      meant to notify about.
      
      Fix this by first updating internal thin-pool state to reflect what the
      DM event is being issued about.  This fixes a long-standing racey/buggy
      userspace device-mapper-test-suite 'resize_io' test that would get an
      event but not find the state it was looking for -- so it would just go
      on to hang because no other events caused the test to reevaluate the
      thin-pool's state.
      
      Cc: stable@vger.kernel.org
      Signed-off-by: default avatarMike Snitzer <snitzer@redhat.com>
      f6c36758